PROTEIN STRUCTURE: INSTRUCTIONS FOR USE Luciana Esposito Istituto di Biostrutture e Bioimmagini Consiglio Nazionale delle Ricerche Via Mezzocannone 16, Napoli E-mail: [email protected] Naples, Italy 29 Aug – 3 Sep 2019 Outline • Sequence – Structure – Function Paradigm • Important information in a PDB entry that can be relevant for a correct use of a protein structure • The relevance of the knowledge of a 3D structure in Drug Design, Molecular Docking, Molecular Dynamics simulations • From 3D structure to function: the complement of sequence to function relationships • Brief outline of Integrative Structural Biology The centrality of Structural Biology Biology - biomolecules Structural Biology Integrative Structural Biology Structural Biology aims to understand how biology works at the molecular level. Structural biology is the study of the molecular structure and dynamics of biological macromolecules, and how alterations in their structures affect their function. Paradigm Sequence The aminoacid sequence encodes the structure Structure The structure determines the function Function The Structure – Function relationship Function Structure Structure Function • The classical way • Post-genomic • A function is discovered and • A new, uncharacterized gene is studied found in a genome • The gene responsible for the • Predictions or high-throughput function is identified methods select this gene for • Product of this gene is further studies isolated, crystallized and the • The protein is expressed and has structure solved to be studied in detail • The structure is used to • The structure is solved and can “rationalize” the function and be the first experimental provide molecular details information about the “hypothetical” protein whose function is unknown Summary of Information Derived from 3D-structure Thornton JM et al., Nat. Struct. Biol. 2000 wwwPDB Since 1971, the Protein Data Bank archive (PDB) has served as the single repository of information about the 3D structures of proteins, nucleic acids, and complex assemblies. Let’s analyze important information that can be retrieved from a PDB entry and that have to be consider when using the structure to address functional questions and/or expand the understanding of the protein system by means of other tools: Drug design Docking MD simulations … Resolution mmCIF 1a6m 106m Resolution is a measure of the level of detail present in the diffraction pattern and the level of detail that will be seen in the electron density map. 108m 1s0h The need of Ultra-high Resolution very subtle details The molecular basis of phosphate discrimination in arsenate-rich environments Periplasmic phosphate-binding protein PBP: arsenate-bound and phosphate-bound structures determined at 0.96 Å and 0.88 Å resolution. The low-barrier Hydrogen Bond (negative- charge-assisted HB) angles are optimal in the phosphate-bound structure but distorted with arsenate. This is the consequence of the longer As-O2 bond than the P-O2 bond. Anion selectivity (at least 103 excess) The energy of this short bond is channelled not towards anion-binding affinity but rather towards anion selectivity Elias M et al. Nature, 2012 Atomic displacement parameters (B-factors) x,y,z coordinates Occ. B-fact mmCIF Atomic displacement parameters (B-factors) Color scheme: Blu Red Low High B B-factors / Missing residues mmCIF Color scheme: Blu Red Low High B Biological Assembly The biological assembly is the macromolecular assembly that has either been shown to be or is believed to be the functional form of the molecule. mmCIF Biological Assembly The biological assembly is the macromolecular assembly that has either been shown to be or is believed to be the functional form of the molecule. Hemoglobin A crystal asymmetric unit may contain: •one biological assembly •a portion of a biological assembly In all structures of Hemoglobin, the •multiple biological assemblies biological assembly is a tetramer. Interfaces in the lattice are biologically relevant? Interface classification problem. Given (a) as a crystal lattice, you could choose any of (b–d) as the biological unit. Further information is needed to identify which arrangement represents the true biological unit. Capitani et al., Bioinformatics 2016 How to distinguish biologically relevant contacts from crystal packing contacts? 1. Evaluate all protein contacts (interfaces) in crystal 2. Leave only the strongest ones - what you get will have major chances to be a stable protein complex (“biologically relevant”) . Method to evaluate the chemical stability of protein assemblies https://www.ebi.ac.uk/pdbe/pisa/ PISA: based on the binding energy of the interface and the entropy change due to complex formation. In the binding energy term: Buried Surface Area (BSA), defined as the difference in Accessible Surface Area (ASA) between uncomplexed and complexed structures; hydrogen bonds, salt bridges and disulfide bonds. 80-90% success rate ; classification is hard in the low range of interface area 400-2000 Å2 – weak biological interfaces more similar to crystal contacts EPPIC: based on geometrical features of the interface but also on evolutionary features (conservation of residues at the interfaces by multiple sequence align. of all sequence homologs) Improving current methods: MD simulations of the interface of interesthttp://www.eppic-web.org Example of importance of crystal packing for information about the biological assembly Convulxin (CVX) is a C-type lectin-like protein from the venom of the South American rattlesnake that functions as a potent agonist of the platelet collagen receptor glycoprotein VI (GPVI). Crystal symmetry 1UMR Dimer (α4β4)2 ASU: α β 2 dimers αβ 4 4 Disulfide linked heterodimer generated by PISA software By using Analytical Ultra-Centrifugation it has been shown that CVX exists in solution as a dimer of α4β4 rings, yielding eight potential binding sites for GPVI. Reanalysis of previously determined crystal structures of CVX revealed the dimer in the available structures. Horii, Brooks & Herr, Biochemistry, 2009 Example of importance of crystal packing for information about the biological assembly PISA software output: Assembly 16-mer Difficult cases: small protein-protein interface Human cytosolic sulfotransferases are classified as monomeric in the PDB structures. PISA and EPPIC failed to find a dimeric stable assembly. It has been confirmed experimentally that the biological unit is a dimer which is present in the crystal structure. Further support from the conservation of the interface across different 2H8K crystal forms of compared crystal of homologous proteins Site-direced mutagenesis of residues at the interface and gel filtration analysis Petrotchenko EV et al, FEBS Letters, 2001 Difficult cases: small protein-protein interface Human cytosolic sulfotransferases are classified as monomeric in the PDB structures. PISA and EPPIC failed to find a dimeric stable assembly. It has been confirmed experimentally that the biological unit is a dimer which is present in the crystal structure. Further support from the conservation of the interface across different crystal forms of compared crystal of homologous proteins http://dunbrack2.fccc.edu/protcid/ Site binding sites (author and/or software annotation) mmCIF NADH 1R37 … … Drug Design The starting point in the development of a new drug is the search for an appropriate lead structure for a target protein. High-throughput screening assay using natural and synthetic product libraries (“random approach”) A “rational” approach involves the computer aided drug discovery. Structure-based route to drug discovery, where the decisive prerequisite is the knowledge about the structure of the target protein. • G. Klebe, Drug Design - The 3D structure is the starting point for the initial Methodology, design of a ligand, which is then to be synthesized Concepts, and Mode- and tested. In the case of good binding, an attempt of-Action, 2013, Springer ed. is made to determine the 3D structure of the protein– • Lu et al, Computer- ligand complex with the new compound. This is the Aided Drug Discovery starting point of the next design cycle. in Accelerated Path to Cures, 2018, Springer Optimization of the lead structure (potency, ed. selectivity, and metabolism) Drug Design: Contribution of protein structures and PDB to New Drug approvals US FDA approved 210 new molecular entities (NME; new drugs) in 2010-2016. A) Small chemicals of MW<1000 Da (81.4%) B) Proteins and peptide hormones with MW>=1000 (17.1%) C) Nucleic acid drugs (antisense oligonucleotides that target mRNA) (1.5%) About 93% of NME have a «relevant» structure in the PDB. ‘Relevant’: Structures that contain a known (>=95% seq id to the Uniprot reference sequence for the target protein) protein target of a NME or one of the NMEs. 95% of relevant structure were determined using X-ray (median resolution limit, 2.08Å). For most of the known protein target of a given NME, the PDB contains more that one relevant structure (80%). A considerable fraction of the relevant structures (37%) include both the NME protein target and one or more partner proteins. Westbrook & Burley, Structure 2019 Structure-based Drug Design In structure-based drug design, attempts are made to design small-molecule ligands by docking them directly into the binding pocket of a target protein. Structure-based design starts with a detailed analysis of the binding pocket to elucidate hot spots for putative interactions with the protein. Either experimental methods or computational